Atypical reciprocal social interaction is a core diagnostic feature of autism spectrum disorder (ASD), but the brain network dynamics contributing to difficulties with social interactions remain unclear. Dominant theories suggest a role for social-motivational and social-cognitive brain networks. However, these theories have been tested in neuroimaging contexts divorced from social interaction (i.e., viewing photos of strangers). Because these detached, offline measures fail to capture the real-world social-interactive challenges faced by individuals with ASD, the critical question of what neural mechanisms underlie atypical social interaction remains unanswered. The current proposal takes an innovative, interactive approach by examining real-time social interaction during fMRI data acquisition to characterize the development and interaction of motivational (i.e., reward) and social-cognitive brain networks during middle childhood. Middle childhood presents an opportune time to address this question because this age is marked by increased social competence in neurotypical (NT) children but plateauing social abilities in ASD. At the neural level, middle childhood is a time of significant change in social brain networks in NT but relatively little data speak to these changes in ASD. Our central hypothesis is that atypical organization of social-motivational and social-cognitive networks during real- time social interactions predict social-interactive difficulties in autism. We will test this hypothesis with the following specific aims in NT and ASD children: 1) determine whether approval from a real-time social partner modulates motivational systems and 2) determine whether real-time social interaction modulates social- cognitive systems, and 3) determine how interactions between motivational and social-cognitive networks during real-time social interaction relate to measures of real-world social behaviors. To achieve these aims, 70 NT and 54 ASD children (7-12 years of age) will participate in a social reward and a social cognitive task in which they engage with an experimenter in real time during functional MRI data acquisition. Graph theoretic measures will be used to examine network organization and interactions within motivational and social- cognitive brain networks during the social interaction task and at rest. Successful completion of the proposed research will advance our understanding of how reward and social-cognitive networks are modulated by real- time social interaction in children with ASD and whether atypical modulation of these networks is related to real-world social impairments. These contributions will be significant to the scientific field and relevant to the strategic mission of NIMH because they will 1) provide critical missing information about the development of brain mechanisms underlying social interaction, 2) determine how individual differences in these brain circuits map onto directly observable behavior, and 3) examine differences in the course and mechanism of functional brain development between neurotypical and ASD children, which will address questions of how, when, and where to intervene.